Before asking NCASE about Micro-ATX...

[Vittra]

X99 on mATX is still limited to 4 DIMMs instead of 8, so not all of the advantages carry over. Overall though, mATX compatibility on a scaled down itx design is, quite like the mATX with ATX, rife with possibilities.

2x Single Slot GPUs (R9 Fury X / Nano) or a single pcb dual gpu (Fury X2) with an add-in card. WC could potentially still be an option, but would be dependent on the actual spacing between the mATX board and the intake fans. Based on the ITX sketch with a rad and attached small res/d5, an matx + slim rad + 25mm fans would fit - if not the 27.5mm Darkside Slim, then likely the 22mm Silverstone Slim. Whether or not such a loop is wise in terms of temps is another story.

Note that in the above mentioned 2x card single slot configuration, this opens the spot between cards - for those with an M.2, it will have some room to breathe, and for those with an actual 1x pci-e slot, it could potentially be used.

 

[rustcoder]

I am wondering if a single 280 radiator is enough to cool a x99 and one or two GPUs?

Any options for
[1] some holes on the roof or the floor for another 240 radiator and fan; or
[2] a pedestal under the case for fan and radiator (similar to Caselabs)?

 

[Necere]

Necere, what are thought about the direction of this discussion? Which internal and external design is your favorite at the moment?

In terms of exterior design I'd be happy to go with the red case (which wouldn't actually be red, to be clear). The side-mounted I/O and lack of optical will no doubt bother some people, but those are kind of intrinsic to the design, and it wouldn't be the same if those were changed. So it may be that we need to just forge ahead, even if those things turn some people off.

In terms of interior, the flexible motherboard support seems like the way to go. The only question is, do we do the larger 5-slot, mATX/ATX 30L case, or the 3-slot ITX/mATX 20L case? This forum will lean towards the latter, I assume. I just wonder if I'll catch a bunch of flak for creating an "almost mATX" case by leaving off the 4th slot. I feel like precluding (for the most part) SLI/Crossfire by design conveys the idea that it's more of an "ITX-plus" case, rather than an mATX case. That is, it's not designed for the cooling and power requirements of SLI/Crossfire, which you really should have 5 slots for (and an ATX PSU, in some cases). But I still feel like there will be people complaining about it.

I am wondering if a single 280 radiator is enough to cool a x99 and one or two GPUs?

Any options for
[1] some holes on the roof or the floor for another 240 radiator and fan; or
[2] a pedestal under the case for fan and radiator (similar to Caselabs)?

I assume you're asking about the 30L concept, since the 20L ITX version only has room for a single 240. The 30L would have a 120mm fan mount at the rear, so 280+120 in total. I don't foresee any options besides those. If you look at other cases in this size class (e.g., TJ08-E), that's about as good as you'll find.

Is it enough? It depends on a lot of factors. How much wattage are you trying to cool? How much noise are you willing to accept to do it? Do you need "low" temps, or are "safe" temps good enough?

 

[vipz]

I think if the side-mounted I/O is made modular and can be placed on left or right and top or bottom, it will minimize the number of people that are bothered. I'm actually warming up to this design.

I'm heavily leaning towards the mATX/ATX version. IMO the ITX version with its SFX PSU limit should be a separate project.

 

[Phuncz]

I think if the side-mounted I/O is made modular and can be placed on left or right and top or bottom, it will minimize the number of people that are bothered. I'm actually warming up to this design.

I'm heavily leaning towards the mATX/ATX version. IMO the ITX version with its SFX PSU limit should be a separate project.

I agree with all these points. In a perfect world, you'd be making both cases at once, catering to most people interested in these. But that's not realistic.

 

[Sverebom]

The side-mounted I/O and lack of optical will no doubt bother some people, but those are kind of intrinsic to the design, and it wouldn't be the same if those were changed. So it may be that we need to just forge ahead, even if those things turn some people off.

Have you considered a solution that allows people to change the position of the I/O? For the I/O it doesn't matter if it is mounted on the left or the right. You "just" need to design side vents that can be turned around.

I agree that getting rid of the ODD would solve many problems. With the 1c layout the ODD would only get in the way of the cooling options and a clean interior, unless you mount it at the very top or near the bottom. Both solutions don't work very well though and/or just don't look very nice.

This forum will lean towards the latter, I assume. I just wonder if I'll catch a bunch of flak for creating an "almost mATX" case by leaving off the 4th slot.

You will. And rightly so I'm afraid. If it is really just about the small extra height for a fourth expansion slot that gets in the way of full mATX-mainboard and ATX-PSU compatibility (let's not forget that ATX-PSUs could almost fit into that case as well), you will get a lot of flak. I think we are only leaning more towards the mITX-/mATX-hybrid because it is small. From a design standpoint the mATX-/ATX-hybrid makes way more sense.

I feel like precluding (for the most part) SLI/Crossfire by design conveys the idea that it's more of an "ITX-plus" case, rather than an mATX case.

In a small ITX-tower I would not go with the variant 1c anyway. In my opinion an ITX-tower should be designed for a small footprint, maybe with the radiator as an exhaust in the roof and the intakes in the bottom side panels and no cooling option in the front panel to save as much room as possible. Since an enthusiast GPU overlaps the mainboard with one to two slots, you have extra room underneath the mainboard that you could probably use for interesting fan setups (like have the fan intakes underneath the mainboard on the right side of the case).


I would buy the "red case" with the clean front panel and the variant 1c internals. However, I'm not convinced that it offers enough over the competiton. Personally I still find the two color "half Mac" design more appealing and unique; the initial design you have posted here as a starting point combined with the perforated front-panel you have posted here and the either the variant 1b internals with the front I/O you have posted here, or the variant 1c internals with the front I/O mounted out of the way of the cooling options. The only problem we might have to solve are the feet. Maybe make them shorter so that they are sturdy enough with just 2mm aluminium.

 

[Vittra]

mATX/ATX is the way to go.

The mITX/mATX can be re-visited in the future once the above case has established itself, and such a point a down-scaled "itx-plus" version that is itx/mATX can be put forth as a smaller version of the bigger design, that offers some more flexibility than the tiny "true-itx" M1.

 

[Necere]

I think if the side-mounted I/O is made modular and can be placed on left or right and top or bottom, it will minimize the number of people that are bothered. I'm actually warming up to this design.

Have you considered a solution that allows people to change the position of the I/O? For the I/O it doesn't matter if it is mounted on the left or the right. You "just" need to design side vents that can be turned around.

It's easier said than done. The I/O is integrated into the vents, and if you just switched sides they'd be upside-down at the bottom. But there may be a way now that I think about it...

You will. And rightly so I'm afraid. If it is really just about the small extra height for a fourth expansion slot that gets in the way of full mATX-mainboard and ATX-PSU compatibility (let's not forget that ATX-PSUs could almost fit into that case as well), you will get a lot of flak.

It would need to be 45mm taller to fit both an mATX motherboard and ATX PSU. At that point you might say, well why not 20mm more for a 5th slot and proper SLI/Crossfire cooling? And then you might say, well, we need a 120mm rear fan to exhaust all that heat. And then we're back to the 30L version.

In a small ITX-tower I would not go with the variant 1c anyway. In my opinion an ITX-tower should be designed for a small footprint, maybe with the radiator as an exhaust in the roof and the intakes in the bottom side panels and no cooling option in the front panel to save as much room as possible.

A top rad is just wasted space if it's not used, though. At least with a front rad the space is useful for longer GPUs and/or drives.

I would buy the "red case" with the clean front panel and the variant 1c internals. However, I'm not convinced that it offers enough over the competiton. Personally I still find the two color "half Mac" design more appealing and unique; the initial design you have posted here as a starting point combined with the perforated front-panel you have posted here and the either the variant 1b internals with the front I/O you have posted here, or the variant 1c internals with the front I/O mounted out of the way of the cooling options. The only problem we might have to solve are the feet. Maybe make them shorter so that they are sturdy enough with just 2mm aluminium.

An idea is to make the foot piece from steel, but then the entire outer shell would need to be painted or powder coated.

 

[Sverebom]

A top rad is just wasted space if it's not used, though. At least with a front rad the space is useful for longer GPUs and/or drives.

That's why I don't like having the radiator in the front because usually I don't use abnormally large graphics card. It's an exception that I'm currently planning with an R9 Fury Strix (because of performance, power consumption, noise, cooling aaaand ... FreeSync). Anyway, I agree that a top-mounted radiator should part of a concept, like very thin case (which is not whithin the scope of this project of course).

An idea is to make the foot piece from steel, but then the entire outer shell would need to be painted or powder coated.

I prefer power coating anyway, and I think that this particular design would look better with it. But Lian Li doesn't do that, right? And I assume that we have no other options.

 

[Necere]

I prefer power coating anyway, and I think that this particular design would look better with it. But Lian Li doesn't do that, right? And I assume that we have no other options.

Not in-house, no. We'd have to outsource it, which would increase cost, though I couldn't say by how much.

 

[vipz]

It's easier said than done. The I/O is integrated into the vents, and if you just switched sides they'd be upside-down at the bottom. But there may be a way now that I think about it...

The way I imagined it when I made the suggestion was to have the vents and the I/O be separates pieces.

An idea is to make the foot piece from steel, but then the entire outer shell would need to be painted or powder coated.

Would it make sense to keep the foot piece as is (2mm aluminum) but support it from the inside with a loop made of steel?

 

[Sverebom]

The way I imagined it when I made the suggestion was to have the vents and the I/O be separates pieces.

Same here. I imagine the vents to be a separate component that you can just rotate around the y-axis. Voilà! All you need now is a clever mounting system that doesn't care about the orientation of the side vents.

As for Lian-Li and powder coating: Do we have other manufacturers to chose from? And are we sure that the feet won't be stury enough, even if you make them smaller and reduce the volume of the case? Do we have other design options for the feet? Could the front an rear panel be tall enough to fill the feet as well?

 

[rustcoder]

or you can have polished stainless steel. No powder coating required.

 

[iFreilicht]

It's easier said than done. The I/O is integrated into the vents, and if you just switched sides they'd be upside-down at the bottom. But there may be a way now that I think about it...

I don't know if we had the same idea, but if that vent-piece was screwed to the case with four screws, one on each corner, couldn't it be made symmetric around the Z-axis?

That way you could have four possible positions for the I/O: left top, left bottom, right top, right bottom.

 

[Phuncz]

I don't know if we had the same idea, but if that vent-piece was screwed to the case with four screws, one on each corner, couldn't it be made symmetric around the Z-axis?

That way you could have four possible positions for the I/O: left top, left bottom, right top, right bottom.

If possible, this should be pursued, that way almost everyone can be happy. And if possible, have a front panel with removable cables with headers, so people can mod these easily (sleeve, extend, shorten, replace cables, etc.).

 

[Urelure]

Have you considered a solution that allows people to change the position of the I/O? For the I/O it doesn't matter if it is mounted on the left or the right. You "just" need to design side vents that can be turned around.

It's easier said than done. The I/O is integrated into the vents, and if you just switched sides they'd be upside-down at the bottom. But there may be a way now that I think about it...

I don't know if we had the same idea, but if that vent-piece was screwed to the case with four screws, one on each corner, couldn't it be made symmetric around the Z-axis?

That way you could have four possible positions for the I/O: left top, left bottom, right top, right bottom.

I think this could be a big deal, so that people aren't limited to putting the case on their right side. This is usually why consumers and reviewers don't like side mounted I/O.

You will. And rightly so I'm afraid. If it is really just about the small extra height for a fourth expansion slot that gets in the way of full mATX-mainboard and ATX-PSU compatibility (let's not forget that ATX-PSUs could almost fit into that case as well), you will get a lot of flak. I think we are only leaning more towards the mITX-/mATX-hybrid because it is small. From a design standpoint the mATX-/ATX-hybrid makes way more sense.

Quite a bold statement. I don't see why offering the possibility for more ram and maybe a pci SSD is something to critique. It is supposed to be an itx case with some expanded functionality. You're basically saying, screw you for letting me put a matx board in my itx case, because what i really want is a matx case in the first place.

It would need to be 45mm taller to fit both an mATX motherboard and ATX PSU. At that point you might say, well why not 20mm more for a 5th slot and proper SLI/Crossfire cooling? And then you might say, well, we need a 120mm rear fan to exhaust all that heat. And then we're back to the 30L version.
In a small ITX-tower I would not go with the variant 1c anyway. In my opinion an ITX-tower should be designed for a small footprint, maybe with the radiator as an exhaust in the roof and the intakes in the bottom side panels and no cooling option in the front panel to save as much room as possible. Since an enthusiast GPU overlaps the mainboard with one to two slots, you have extra room underneath the mainboard that you could probably use for interesting fan setups (like have the fan intakes underneath the mainboard on the right side of the case).

Exactly

 

[jalex3]

Shame more people aren't interested in top mounted PSU layouts. Not sure why that is.

 

[psurge]

The mac/2 design with perforated black front panel sounds great to me. Stainless steel is an idea, but would it actually be cheaper than thick aluminum? Isn't it significantly harder to work with than aluminum? Also, I'm not sure Necere's heart is in that design...

For the red case: I don't know. Speaking solely aesthetically, it's somehow either too minimal for me or not minimal enough, if that makes any sense. There's so little going on that the vents, IO, and panel gaps really stand out, and I find myself wanting them to go away, leaving a space odyssey style metallic monolith. Maybe it's just the red that's really making all the other stuff pop too much though .

 

[iFreilicht]

Shame more people aren't interested in top mounted PSU layouts. Not sure why that is.

What are the benefits of such a layout? What does it offer that a bottom mounted PSU can't?

 

[Aibohphobia]

The PSU isn't blocking the video card mainly.

 

[Necere]

The way I imagined it when I made the suggestion was to have the vents and the I/O be separates pieces.

It's more complicated if you want to maintain a clean appearance and add some vertical rigidity to the panel.

Would it make sense to keep the foot piece as is (2mm aluminum) but support it from the inside with a loop made of steel?

I think that would be rather ugly.

As for Lian-Li and powder coating: Do we have other manufacturers to chose from?

We haven't found a suitable alternative at this point.

And are we sure that the feet won't be stury enough, even if you make them smaller and reduce the volume of the case?

Volume doesn't necessarily correlate to weight. The weight of the components is far greater than the weight of an aluminum case anyway, so if making it smaller means making it lighter, then you're talking about cutting features.

Do we have other design options for the feet? Could the front an rear panel be tall enough to fill the feet as well?

It wouldn't really be the same then, would it? I think much of the appeal of the "hoops" is that they are hoops, and visually separated from the case proper.

or you can have polished stainless steel. No powder coating required.

Making the lower section from steel while keeping the rest aluminum would not be a good look.

I don't know if we had the same idea, but if that vent-piece was screwed to the case with four screws, one on each corner, couldn't it be made symmetric around the Z-axis?

That way you could have four possible positions for the I/O: left top, left bottom, right top, right bottom.

Keep in mind the audio jacks will have decals that are orientation-dependent. Also I don't want to use visible screws if it can be helped.

Shame more people aren't interested in top mounted PSU layouts. Not sure why that is.

A bottom-mounted PSU lets it have a dedicated air intake, lowers the center of gravity of the case, makes cable management a bit easier (they'll sit in the bottom instead of hanging down).

What are the benefits of such a layout? What does it offer that a bottom mounted PSU can't?

One of the benefits for a compact case, specifically, is that it allows you to have a 5.25" optical drive in the "normal" top location without it running into the motherboard. The TJ08-E is a good example of that, where if you were to flip it so the PSU is on the bottom, the drives would have to go with it, or else the case would need to be longer to avoid fouling the motherboard.

 

[Phuncz]

About the front vents and flexible I/O block location mounting:

Keep in mind the audio jacks will have decals that are orientation-dependent. Also I don't want to use visible screws if it can be helped.

I hope and believe you could make it work. If the I/O block is seperately removable, the orientation isn't going to be an issue as you could keep it the same orientation for all four corners. The screws for the I/O block could be done from behind to hide them, the ones for the grill could be possibly placed horizontally to hide the top one below the panel and the bottom one inside the case. Ofcourse this is all considered in a vacuum as we don't know the internal design yet.

You might have already considered this, I just wanted to add my idea. But I think for that design it's one of the most important features to get perfectly since it's pretty featureless due to it's simplicity.

 

[Urelure]

If all else fails regarding swappable I/O, if you could flip the entire front panel assembly up side down as a last resort, at least anyone really needing the I/O on the other side could, if they were willing to live with the decals and power symbol being up side down as well.
All you would need for this to be possible is for the pegs or screws (or whatever is used to attach it) line up both ways. It wouldn't take away from the design as intended, and might make the side I/O less of a deal breaker for some.

 

[10e]

I think the Silverstone TJ08-E layout, as well as that used in the KL06B is excellent.

My only issue with that case is that it only has four PCI-E backplates effectively eliminating the use of SLi or Crossfire with motherboards that support the extra PCI-E x8/x8 spacing that Gigabyte popularized a while back and is still in use with the Asus Z97M-Plus and MSI Z97M-G43.

To keep a cleaner top appearance, flipping the PSU so the intake fan is at the bottom would work wonderfully to maintain the nCase solid top panel aesthetic and give some extra exhaust oomph under load (with those PSUs that have the fan that only starts when needed for cooling). In terms of intake and not getting heated by GPUs below it, you could add 5mm to height and put an intake/diverter plate to receive air more from front facing 120/140mm intake fans. Axial fans outside the Air Penetrators have some "spray" or cone shaped airflow patterns which would work fine here to add cool front intake. above their plane of air into the PSU. In this case, low, or bottom mounted I/O with appropriate cable management cutouts would also work well for those of us who like compact desktop located PCs.

As for powder coated steel versus anodized aluminum, powder coated steel looks great (to me) but anodized aluminum has two advantages: Easy to paint due to the porous surface accepting primer and paint very well in terms of adhesion, and is generally more rigid and doesn't require the extra rigidity bends/folds that steel needs to avoid flex.

Powder coating has to be stripped to paint as well. I've been there, failed that

I don't think you necessarily have to stick to the 312mm afforded for PCI-E cards, I think you can shave off a few millimetres in the name of compact dimensions and go down to 295-300mm to save space. Very few video cards exceed 300mm in length, even dual GPU versions. This would allow a set of 25mm front intake fans and potentially a 30-35mm radiator to be attached to the front.

In addition, using this layout allows you to take advantage of bottom intake for a 240mm radiator, and front covered intake for a balance of silence versus airflow for a 240mm/280mm front radiator. You can eliminate extra space usage by off-setting the bottom intake fans similarly to what Fractal Design has done with the Arc Mini R2 and Define R5. The Arc Mini R2 is only 210mm wide, which is still relatively small.

You can then set up a triple 2.5/3.5" disk cage in the front for mechanicals/SSDs between the top mounted PSU and a front optical slim drive mounted vertically that can have the access slot up top similar to the NCase you currently sell. I'd prefer this approach to Silverstone's steadfast and stubborn approach to having space allocated for a full sized optical up top. I barely use opticals and slims work just fine for me.

I also like the reverse M-ATX layout because I'm a steadfast watercooler and it allows me to put my reservoir below my video cards, effectively using the open space below GPUs more efficiently like I did in the Raijintek Metis last year with XSPC RX120 V3 radiator:

https://dl.dropboxusercontent.com/u/21422895/Lil%20Bit/2015-01-24%2005.47.47.jpg

I haven't done any calculations in overall height/width/length (or depth) on this type of setup but I think you can fit it into your desired "displacement" target, maintain a clean and subtle appearance and still have great cooling balanced with silence.

Just my $0.02. I think you are doing a great job, Necere. Oh yeah and that other guy too

 

[SaperPL]

Woah, didn't read for a bit and whole thread changed into waterfall of text

TL;DR,

As for the concepts and ideas: I don't think that you can make something really interesting unless you go for the psu over cpu setting or vertically in front.

Putting psu in normal config, in my opinion, just makes it a generally bad idea since there's tons of cases to pick from and unless you're making a nice SFF that stands on the desk, the size doesn't matter.

There's also option of making something like chx proposed here: http://hardforum.com/showthread.php?t=1878878

Make a mATX/ATX case with SFX psu over slots that will support only one card in primary slot. I think going for something like that would be cool because of more memory support and general availability/pricing of the boards compared to mITX.

 

[Urelure]

To keep a cleaner top appearance, flipping the PSU so the intake fan is at the bottom would work wonderfully to maintain the nCase solid top panel aesthetic and give some extra exhaust oomph under load (with those PSUs that have the fan that only starts when needed for cooling).

The main issue with this is that the PSU and GPU, if SLI/CF, would be competing for air. The typical scenarios where they are under load and require extra cooling are in most cases identical, which means that they would most likely reduce each others cooling capacity when they need it the most. Since you say you are a steadfast watercooler this of course would be less of an issue for you, but it's still very much an issue imo.

I don't think you necessarily have to stick to the 312mm afforded for PCI-E cards, I think you can shave off a few millimeters in the name of compact dimensions and go down to 295-300mm to save space. Very few video cards exceed 300mm in length, even dual GPU versions. This would allow a set of 25mm front intake fans and potentially a 30-35mm radiator to be attached to the front.

If you look at this post http://hardforum.com/showpost.php?p=1041910522&postcount=144 necere looks to be leaning towards this exact thing. leaving the depth to a point where you have to sacrifice some GPU length for the 240/280 rad in the front.

As for the concepts and ideas: I don't think that you can make something really interesting unless you go for the psu over cpu setting or vertically in front.

Putting psu in normal config, in my opinion, just makes it a generally bad idea since there's tons of cases to pick from and unless you're making a nice SFF that stands on the desk, the size doesn't matter.

This is of course just throwing opinions around, but I don't think it is necessary to do something differently just for the sake of different. I think the limitations introduced by your suggestions would put too many people off it compared to who get won over.

I don't think people will buy this because it offers functionality and/or innovation previously unavailable, but rather because it is really really nice.

 

[Vittra]

The main issue with this is that the PSU and GPU, if SLI/CF, would be competing for air. The typical scenarios where they are under load and require extra cooling are in most cases identical, which means that they would most likely reduce each others cooling capacity when they need it the most. Since you say you are a steadfast watercooler this of course would be less of an issue for you, but it's still very much an issue imo.

Keep in mind the following:

1) We are referring to ATX PSUs, most of which do not employ the use of a fan unless a certain temperature threshold is met. It is likely that even with an axial fan sitting right below it spewing hot air towards it, the intakes providing fresh air to move that warmer air negates it sufficiently that said threshold will not be a concern, which means there will be no fighting occurring.

The scenario outlined above is typically what occurs in the TJ08-E when the PSU's intake fan is placed to intake from inside the case.

2) The post you are quoting does mention the addition of a diverter plate for the PSU if necessary, which would thermally isolate the two sections.

I don't think people will buy this because it offers functionality and/or innovation previously unavailable, but rather because it is really really nice.

I disagree. People will not buy this over a design that already exists and is cheaper, simply because it is nicer. It does need to provide something beyond that. Or, perhaps I should say, it will not be bought in that volume that Necere/Wahaha360 need to consider to meet MOQ.

 

[SaperPL]

This is of course just throwing opinions around, but I don't think it is necessary to do something differently just for the sake of different. I think the limitations introduced by your suggestions would put too many people off it compared to who get won over.

I don't think people will buy this because it offers functionality and/or innovation previously unavailable, but rather because it is really really nice.

I don't feel the idea of making something look really really nice but internally just slightly different than mainstream cases. I'm not sure but I think there's going to be a ton of cases to choose from and argument of being slightly looking better since it won't use plastic doesn't get to me. I might be biased and that's just my opinion though.

I think that it's crucial even for the design team to set a goal to be something cool and innovative or they'll burn out quickly spending their time making the same thing that was already mass produced.

If I were to build such thing, I'd definitely want to make it unique in functionality and size. Making a SFF case supporting amd octa cores would be cool for a start since there's always this question if there's anything mATX/mITX that supports them properly. And since AMD is preparing their next arch to hit soon-ish there might be another of such power hungry but over 9000 in terms of cores to think of.

 

[Sverebom]

This is of course just throwing opinions around, but I don't think it is necessary to do something differently just for the sake of different. I think the limitations introduced by your suggestions would put too many people off it compared to who get won over.

But what is the selling point then if people could get a similar case from manufacturers like Silverstone for probably less than what an mATX Ncase would cost? A clean front panel without any functionality? Is that the reason why people should pay 200+ Dollars for a fairly common internal layout (that it less capable than every $80 case)?

Having said that, placing the PSU over the CPU socket isn't exactly new and unique either. My first Lian Li case had the same layout 15 years ago. A Silverstone SG10 with a clean exterior and a more sensible cooling design would be something. And it would be a smaller and less complex build, which in return could help to keep the price low, and could look like a larger M1 (an external design that has already found a market!!!) and thus might be the logical step forward from the M1. But - "Yuck!" - it's soooo wide, right?

Or the Variant 1b with the PSU and the drives - inluding the ODD - in the top of the case for a clean divison between the component section and the drives section. No interference between the sections and a clean airflow with just three fans. While actually a very old school internal layout, it might have become fresh and interesting again in market that has gone nuts with the amount of fans and radiatiors that you can cram even into "small" cases these days. With a Mac-style exteror it could be a winner.

Or what about this: A case that is as slim as possible with native support only for standard width expansion cards, probably not wider than 160mm without space for cable management behind the mainboard. The clou: The rear panel could be modular so that oversize graphics card can be mounted upright in front of the mainboard with the help of a PCIE extension (which shouldn't be included). That probably won't allow a multi-GPU setup with oversize graphics cards, but with that form-factor and with that little room between the graphics card it would be better to use blower-style reference cards anyway.

Furthermore that case would have a bottom-to-top airflow with a radiator in the top. Since full-size tower coolers won't fit into the case, the necessary space above the mainboard wouldn't be wasted because AiO-watercoolers would be the go-to solution for best cooling performance. Only problem: Where to put the PSU and the drives? Maybe in that space next to the CPU- and RAM-area.

Most likely a stupid concept that wouldn't work. But at least it is something different and unique, like the M1. Maybe we should start to think in directions that have haven't been explored yet.

 

[3leven]

I'm a total noob at this so go easy on me. I like the idea of mATX/ATX and played around with combining the idea with 2.d so give me your thoughts on this.

Instead of the PSU overlapping the ATX board, you can save some height by moving it to the top/front of the case, and move the ODD from above the PSU in 2.d
We aren't looking to fully support ATX boards anyway, so having it go to the bottom of the case "prevents" the use of tri-sli configs without having to overlap. When using an ATX board, the drawbacks would be losing space for a 240 rad and more storage as you'll see later.

When looking at using a mATX board, you get a bunch of options like a 240mm rad on the bottom of the case or more HDD bays. Even though we removed the ODD from the top of the PSU, it could be brought back as a modular option behind the PSU.

My rough estimate of the internals is 325x365 and being 210mm wide (as Necere said would be needed with a rotated psu) the case would be around ~26L.
+no top ventilation
+uses the awesome 'red case' design with side vents (and the duct for the psu)
-wide
-who needs front panel i/o anyway?

Maybe we should start to think in directions that have haven't been explored yet.

 

[rustcoder]

Interesting. For those that do not need 3.5" hard drive or slim odd, but want more radiator,

remove the pink hd and odd (on top and sandwiched between the mobo on the left and the power supply on the right), move the orange power supply to the left and place a radiator on the right, in front of the power supply. The fan of the power supply can be fan #1 of the radiator, with a #2 radiator fan below the power supply..

 

[Necere]

I'm a total noob at this so go easy on me. I like the idea of mATX/ATX and played around with combining the idea with 2.d so give me your thoughts on this.

It's not a bad take on the SG09 layout. The front ports are a little problematic on the red case design, since the PSU exhausts right where they would be, and it's a wide layout. But at least with that width tall CPU coolers/GPUs and cable management wouldn't be a concern. The possibility of tri-SLI is interesting, though the PSU length is a limiting factor there. I'll have to explore the idea further.


By the way, I've added the M1 layout to my design post, along with a further explanation of why it doesn't make a ton of sense for mATX. People keep bringing it up, so I figured it needed a dedicated section with pictures for clarity.

 

[Vittra]

It may have the same issue as 2c however, whereby the CPU cooler will be starved for air with 2x GPU - if one were to go for air that is.

 

[psurge]

For the red case - to make the side front io ports relocateable, would it work to have fixed side vents and power button, with the USB ports and jacks on a board that screws into the back of the vent pieces? You'd insert cables through a vent hole, which might make for an even cleaner look. Or it might look terrible and/or make the ports themselves hard to use. I don't know, just throwing it out there.

 

[Urelure]

1) We are referring to ATX PSUs, most of which do not employ the use of a fan unless a certain temperature threshold is met. It is likely that even with an axial fan sitting right below it spewing hot air towards it, the intakes providing fresh air to move that warmer air negates it sufficiently that said threshold will not be a concern, which means there will be no fighting occurring.

I get that it's ATX. Thing is. If a PSU fan and a GPU fan is facing each other. They would both be trying to suck air from the same thin gap between them. The GPU would not blow air into the PSU. And if you are stressing two GPU's, you're probably stressing the PSU as well unless you over-spec the PSU. I don't think the moving air from the GPU fans will be enough to cool the PSU at that point, but I can't prove it

2) The post you are quoting does mention the addition of a diverter plate for the PSU if necessary, which would thermally isolate the two sections.

Yes. But my point was that 10e's reasoning behind trade offs in case depth seems to be very much in line with Necere's own preferences.

I disagree. People will not buy this over a design that already exists and is cheaper, simply because it is nicer. It does need to provide something beyond that. Or, perhaps I should say, it will not be bought in that volume that Necere/Wahaha360 need to consider to meet MOQ.

I don't feel the idea of making something look really really nice but internally just slightly different than mainstream cases. I'm not sure but I think there's going to be a ton of cases to choose from and argument of being slightly looking better since it won't use plastic doesn't get to me. I might be biased and that's just my opinion though.

But what is the selling point then if people could get a similar case from manufacturers like Silverstone for probably less than what an mATX Ncase would cost? A clean front panel without any functionality? Is that the reason why people should pay 200+ Dollars for a fairly common internal layout (that it less capable than every $80 case)?

I am not sure we'll ever agree on this. It's not like I can say that you are wrong and I am right. Out of those that are active in this discussion I seem to be outnumbered though

If the scope of this project was to make a true SFF mATX case (which from what I understand, its not) then you could, for the sake of size, make all sorts of compromises which could lead to an innovative and/or unique case. But if size is not the main goal (Necere has already stated its not) then you're left with making a case which offer as wide an application as possible, with a design that hits home with as many people as possible. I guess its a matter of deciding whether we want this discussion to get the best result for the scope as it is today, or if we think the scope itself should change.

There is a reason why traditional internal layouts has become the norm. No matter how boring, it really works. Either way. The possibility to cram an ATX board in there is kind of unique, so there's always that.

If I were to build such thing, I'd definitely want to make it unique in functionality and size. Making a SFF case supporting amd octa cores would be cool for a start since there's always this question if there's anything mATX/mITX that supports them properly. And since AMD is preparing their next arch to hit soon-ish there might be another of such power hungry but over 9000 in terms of cores to think of.

There is probably something here I am missing, but do you mean mATX with more extensive cooling options due to the high TDP of these chips? wouldn't a 240 rad suffice for this? So shifting the focus to sub 20L and 240 rad support?

 

[Vittra]

No movement in this thread for awhile!

I am not sure we'll ever agree on this. It's not like I can say that you are wrong and I am right. Out of those that are active in this discussion I seem to be outnumbered though

If the scope of this project was to make a true SFF mATX case (which from what I understand, its not) then you could, for the sake of size, make all sorts of compromises which could lead to an innovative and/or unique case. But if size is not the main goal (Necere has already stated its not) then you're left with making a case which offer as wide an application as possible, with a design that hits home with as many people as possible. I guess its a matter of deciding whether we want this discussion to get the best result for the scope as it is today, or if we think the scope itself should change.

There is a reason why traditional internal layouts has become the norm. No matter how boring, it really works. Either way. The possibility to cram an ATX board in there is kind of unique, so there's always that.

The issue is that the TJ08-E (and it's updated brother - the KL06) already offers a case design with a very wide application of uses, while being smaller than 95% of the other mATX cases on the market... all while being over 4 years old at this point. If the only intent here is to make a more aesthetically appealing / better constructed version of it, there is a real concern of actually selling the volume required to meet a MOQ is a real concern.

I, personally, would be 100% fine with a better constructed / aesthetically pleasing version of the above case. I've owned 2x TJ08-E because, despite it's faults, there's really nothing else in mATX that combines it's dimensions and flexibility - this thread is testament to that.

But are there 999 other people out there who would be willing to pay double the price or more than a TJ08-E / KL06? While I'm speculating, I assume Necere / Wahaha360 can't afford to just keep ~500 of these cases sitting around, slowly being sold. Not if they want to move on other projects too.

There is probably something here I am missing, but do you mean mATX with more extensive cooling options due to the high TDP of these chips? wouldn't a 240 rad suffice for this? So shifting the focus to sub 20L and 240 rad support?

I think that's what SaperPL meant, but he'd have to clarify. However, a sub 20L design is being tackled by another project - the Kimera Industries Nova (17L). It will have 2x 240 rad support, depending on component choices. It's already been suggested that Necere isn't interested in competing against Aibo's design nor is this the original objective of focus, though.

From a perspective of cooling, a 5960X + 2x GPU even without any overclocking, while absolutely possible, would introduce some serious compromise - either in temps or acoustics.

 

[SaperPL]

What I meant is that top AMD chips run properly only on full ATX boards, there were hardly any boards to support 125W 8core FX's. Add 4 or even 8 sticks of ram to that and making a ATX case that's not for the sake of SLI make sense again.

It's just that with ATX there's always some stuff you won't get in mATX/mITX format. For example there are boards now with two m.2 slots, I think. What about that?

I watched a lot of different case reviews since I started my case project to know what's out there now and the idea of making mATX case that will support huge water cooling kits with power supply in normal config just isn't worth doing as such project with small MOQ and high price. I get it that it'll be looking pretty nice but for a case standing on the ground under the desk it doesn't matter as well as this tiny difference in volume vs high end mass produced cases. it's simple as that - there's competition already on the store shelf.

 

[3leven]

It keeps going back to what is going to be worth making.

I think we should add 240 rad support (could have drawbacks like gpu length or number of pci slots you can use) and ATX board support to the design requirements officially.

We've bounced enough ideas around where it seems difficult to make a case exclusive for mATX boards that will be worth producing (within our old constraints). And when have you heard of a mATX or ATX case without 240 rad support?

I think it comes down to airflow and psu location now. With ATX support, I don't think we can have the PSU above the board like in the TJ08-E, it will be too tall and space inefficient. Our other options are in the front of the case at either top or bottom, or overlapping the PSU and the board like in the original post.
Also, reversing the board position for a Prodigy M type layout would create a ton of dead space at the top if not using an ATX board.

I think the key here is either going with something that is radical/untried, or really going after either the SG09 or TJ08-E design and refining it.

The TJ08-E design aside for above reasons, I still think your statement hold weight. The question is only adding ATX support going to be radical/untried enough to produce? Per my other post, I'm interested in seeing where the SG09 can be improved.

 

[SaperPL]

I think that good idea might be a combo of two ideas:

1) ATX with psu over last 3 slots proposed here http://hardforum.com/showthread.php?t=1878878

2) ATX with psu over the cpu.

Make the PSU mountable in both positions. Add mounts for water cooling rads in front of the case.

This way you either have "lower end" build that won't obstruct the cpu cooler but have one or two cards max or you use cpu water cooling and put the PSU over it to have all slots available for gpu's.

Making the dual mount for either card or PSU over the motherboard card area might be tricky though.

 

[3leven]

I think that good idea might be a combo of two ideas:

1) ATX with psu over last 3 slots proposed here http://hardforum.com/showthread.php?t=1878878

2) ATX with psu over the cpu.

Make the PSU mountable in both positions. Add mounts for water cooling rads in front of the case.

This way you either have "lower end" build that won't obstruct the cpu cooler but have one or two cards max or you use cpu water cooling and put the PSU over it to have all slots available for gpu's.

Making the dual mount for either card or PSU over the motherboard card area might be tricky though.

With the PSU over the CPU design, do you intend to have all of the expansion slots available? Or still limiting to 4/5?